The reason why you find problems with the non-PDF formats is that it's been OCRed. That is an acronym for Optical-Character-Recognition program, a program that scans and converts pictures to text. It's also called machine transcription.
These programs make text files which are small files full of text, and usually not much else (like pictures).
The OCR program is only a program, not a person, so if it finds text that is smudged or if it glitches you come up with funny unreadable text like that. Sometimes you can interpret what it thought the text was, sometimes the OCR thought scratches and smudges were text, and there wasn't any there, so forget it.
PDFS are normally scanned images, pictures of books, there's no interpretation to text necessary to read them.
This makes them easier to read in many cases. But they're much larger to download because of this, as pictures take up more space than text. Also the picture scans may be a bit off too.
For some e-books, volunteers correct typos and submit the correct copies to e-text archives like Gutenberg, the Internet Archive, Google Books, etc.
That's when you find easy to read clean copy without errors.
All the files uploaded as PDFs to the Internet Archive tend to have an automatic OCR done other formats, often a very poor unreadable one. You may get a readable copy, you may not.
FWIW --
Saints' Books only offers non-PDFs if they've been proofed and typo-corrected. So the books are always readable. I prefer non-PDFs personally, even though the site is full of PDFs and there's always an opening for more volunteers and helpers to help out transcribing and correcting copy. :smile:
Topic: problem we-texts? (Read 838 times)